3D Segmentation of Multi-Contrast Cardiac Magnetic Resonances With Topological Correction and Synthetic Data Augmentation

3D Segmentation of Multi-Contrast Cardiac Magnetic Resonances With Topological Correction and Synthetic Data Augmentation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article 3D Segmentation of Multi-Contrast Cardiac Magnetic Resonances With Topological Correction and Synthetic Data Augmentation Ricardo M Rosales, Manuel Doblaré, Esther Pueyo This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7965351/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Apr, 2026 Read the published version in Annals of Biomedical Engineering → Version 1 posted You are reading this latest preprint version Abstract Purpose: Automatic segmentation of cardiac magnetic resonance (CMR) images improves the evaluation of heart structure and function, helping clinical diagnosis and the generation of in silico models. Recent advances have introduced synthetic augmentation (SA) using generative adversarial networks (GANs) and topological correction (TC) via persistent homology to enhance segmentation with convolutional neural networks (CNNs). However, their combined effectiveness remains unexplored. Here, we extend and systematically evaluate these techniques, individually and in combination, for the first time in the context of three-dimensional (3D) CMR segmentation across challenging multi-vendor, multi-center, multi-class and multi-contrast data sets. Methods: Data involved anisotropic, topologically inconsistent cine and late gadolinium-enhanced (LGE) CMRs, and isotropic, topologically consistent ex vivo CMRs. Topological priors were defined in each data set from ground truth label (GTL) assessments, and TC was applied by retraining the baseline 3D CNN with a loss function accounting for topological discrepancies. For SA, deformed GTLs were used to generate synthetic images using trained 3D GANs. Results: Consistent segmentation improvements were observed for the ex vivo data in both overlap with GTLs and topological precision when applying TC and SA individually and in combination. Notably, an enhanced identification of the infarction was obtained when SA and TC were used in the LGE data. Overall, SA increased the predictions overlap with GTLs, while TC reduced the topological discrepancies across all data sets. Conclusion: TC and SA demonstrate strong potential for improving 3D CMR segmentation on complex, real-world data sets, especially when topologically consistent data are available for training. Biomedical Engineering Artificial Intelligence and Machine Learning Cardiac Magnetic Resonance Convolutional Neural Networks Generative Adversarial Networks Persistent Homology Synthetic Data Augmentation Topological Correction Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Published Journal Publication published 17 Apr, 2026 Read the published version in Annals of Biomedical Engineering → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7965351","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":536015809,"identity":"8e4081c3-9d51-442c-b598-69217e023188","order_by":0,"name":"Ricardo M Rosales","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzElEQVRIiWNgGAWjYNCCAgYGfiB1gAQtBgwMkg0kazEgWr357MOHP/wwsLE3vpF78OCPmnsM/PwENMucS0uT7DFIS9x2Iy/hgMSxYgbJGQn4tUjw8Jgx8BgcTjC7kWNwwIAtgcHgBgGHAbUYf/xj8N/eeAZQS8K/BAb78wQcBtRiIM1jcIBxgwRQy8E2oC0MBB3GliYtY5CcOOPMG4ODjX0JPBI3CGphPvzxTYWdPX97jvHHH98S5Pj7CTgMA/CQqH4UjIJRMApGATYAAE9LPl717y66AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-8867-3802","institution":"Universidad de Zaragoza","correspondingAuthor":true,"prefix":"","firstName":"Ricardo","middleName":"M","lastName":"Rosales","suffix":""},{"id":536015810,"identity":"a677e7d7-d5fb-496a-ad2a-ce51376ed122","order_by":1,"name":"Manuel Doblaré","email":"","orcid":"https://orcid.org/0000-0001-8741-6452","institution":"Universidad de Zaragoza","correspondingAuthor":false,"prefix":"","firstName":"Manuel","middleName":"","lastName":"Doblaré","suffix":""},{"id":536015811,"identity":"392fadf7-3902-40cb-87fe-b905b23c19ad","order_by":2,"name":"Esther Pueyo","email":"","orcid":"https://orcid.org/0000-0002-1960-407X","institution":"Universidad de Zaragoza","correspondingAuthor":false,"prefix":"","firstName":"Esther","middleName":"","lastName":"Pueyo","suffix":""}],"badges":[],"createdAt":"2025-10-28 12:52:00","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7965351/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7965351/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10439-026-04115-8","type":"published","date":"2026-04-18T00:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":107462998,"identity":"e4da5353-2b97-4b2e-8d29-d89498e7b3b0","added_by":"auto","created_at":"2026-04-21 17:35:29","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6261419,"visible":true,"origin":"","legend":"","description":"","filename":"paperCMRRMR.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7965351/v1_covered_a380103c-fc76-4a89-a18c-34bfa5dd95ee.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e3D Segmentation of Multi-Contrast Cardiac Magnetic Resonances With Topological Correction and Synthetic Data Augmentation\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[{"identity":"f5ab78ae-f6a7-46a5-a7b1-ea93578a6b1f","identifier":"10.13039/100010661","name":"Horizon 2020 Framework Programme","awardNumber":"874827","order_by":0},{"identity":"b5a36848-7c7d-455a-adc8-f891382e236e","identifier":"10.13039/501100011033","name":"Agencia Estatal de Investigación","awardNumber":"PID2022-140556OB-I00","order_by":1},{"identity":"8a443c25-0fa5-4f4d-af31-135971d230d9","identifier":"10.13039/501100011033","name":"Agencia Estatal de Investigación","awardNumber":"TED2021-130459B-I00","order_by":2},{"identity":"c73e5d90-f8c6-4844-83e9-74f7bcf06f5d","identifier":"10.13039/501100011033","name":"Agencia Estatal de Investigación","awardNumber":"PLEC2021-008127","order_by":3},{"identity":"2d41d92d-97bc-4984-85b3-3fa343c1e5a2","identifier":"10.13039/501100004895","name":"European Social Fund","awardNumber":"LMP94_21","order_by":4},{"identity":"8b2024a6-acc4-4fdb-b717-754912297948","identifier":"10.13039/501100004895","name":"European Social Fund","awardNumber":"T39_23R","order_by":5},{"identity":"fbea89f8-364f-46da-8760-015d5c8c7837","identifier":"10.13039/501100000781","name":"European Research Council","awardNumber":"638284","order_by":6}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"University of Zaragoza","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Cardiac Magnetic Resonance, Convolutional Neural Networks, Generative Adversarial Networks, Persistent Homology, Synthetic Data Augmentation, Topological Correction","lastPublishedDoi":"10.21203/rs.3.rs-7965351/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7965351/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003ePurpose:\u003c/strong\u003e Automatic segmentation of cardiac magnetic resonance (CMR) images improves the evaluation of heart structure and function, helping clinical diagnosis and the generation of \u003cem\u003ein silico\u003c/em\u003e models. Recent advances have introduced synthetic augmentation (SA) using generative adversarial networks (GANs) and topological correction (TC) via persistent homology to enhance segmentation with convolutional neural networks (CNNs). However, their combined effectiveness remains unexplored. Here, we extend and systematically evaluate these techniques, individually and in combination, for the first time in the context of three-dimensional (3D) CMR segmentation across challenging multi-vendor, multi-center, multi-class and multi-contrast data sets.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e Data involved anisotropic, topologically inconsistent cine and late gadolinium-enhanced (LGE) CMRs, and isotropic, topologically consistent \u003cem\u003eex vivo\u003c/em\u003e CMRs. Topological priors were defined in each data set from ground truth label (GTL) assessments, and TC was applied by retraining the baseline 3D CNN with a loss function accounting for topological discrepancies. For SA, deformed GTLs were used to generate synthetic images using trained 3D GANs.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e Consistent segmentation improvements were observed for the \u003cem\u003eex vivo\u003c/em\u003e data in both overlap with GTLs and topological precision when applying TC and SA individually and in combination. Notably, an enhanced identification of the infarction was obtained when SA and TC were used in the LGE data. Overall, SA increased the predictions overlap with GTLs, while TC reduced the topological discrepancies across all data sets.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion:\u003c/strong\u003e TC and SA demonstrate strong potential for improving 3D CMR segmentation on complex, real-world data sets, especially when topologically consistent data are available for training.\u003c/p\u003e","manuscriptTitle":"3D Segmentation of Multi-Contrast Cardiac Magnetic Resonances With Topological Correction and Synthetic Data Augmentation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-29 00:37:37","doi":"10.21203/rs.3.rs-7965351/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d2b04884-ccfa-4c9e-aca1-a0fd9c94c9e7","owner":[],"postedDate":"October 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":56988908,"name":"Biomedical Engineering"},{"id":56988909,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2026-04-21T17:35:18+00:00","versionOfRecord":{"articleIdentity":"rs-7965351","link":"https://doi.org/10.1007/s10439-026-04115-8","journal":{"identity":"annals-of-biomedical-engineering","isVorOnly":false,"title":"Annals of Biomedical Engineering"},"publishedOn":"2026-04-18 00:00:00","publishedOnDateReadable":"April 18th, 2026"},"versionCreatedAt":"2025-10-29 00:37:37","video":"","vorDoi":"10.1007/s10439-026-04115-8","vorDoiUrl":"https://doi.org/10.1007/s10439-026-04115-8","workflowStages":[]},"version":"v1","identity":"rs-7965351","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7965351","identity":"rs-7965351","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}